A Concise and Modular Synthesis of Pyranicin

Abstract

A modular, 13 step, synthesis of the tetrahydropyran-containing annonaceous acetogenin pyranicin is reported. Key features are the use of an Achmatowicz oxidation-Kishi reduction sequence for the assembly of a pyranone from a furan, and the application of Fu’fs alkyl-alkyl Suzuki coupling for subunit union.

The annonaceous acetogenins, a group of polyketides that are generally identified by the presence of one or two tetrahydrofuran rings, now number greater than 400 compounds.¹ A small subset of the acetogenins, including mucocin and jimenezin, contain both a tetrahydrofuran ring and a tetrahydropyran ring. As a class, these compounds have attracted substantial scientific attention due to their rich and varied biological activities which include anticancer, antiinfective, immunosuppressive, pesticidal, and antifeedant properties.

Pyranicin (1, Figure 1) was isolated from the stem bark of Goniothalamus giganteus Hook. f. & Thomas (Annonaceae) and is the prototype of acetogenins that contain a single tetrahydropyran ring.² Despite what could be argued to be substantial structural simplification relative to more complex acetogenins, pyranicin retains impressive biological activity, with <1μg/mL ED₅₀ values against a number of cancer cell lines and particularly noteworthy effects against the PACA-2 (pancreatic cancer) cell line (ED₅₀ = 1.3 ng/mL). As part of our ongoing interest in the synthesis of natural product inhibitors of electron transport,³ we describe in this communication a total synthesis of 1.⁴

Figure 1.

Pyranicin and overall synthesis strategy.

From the outset, an overarching goal of our studies was to develop a concise and modular synthesis that would be amenable to the synthesis of a variety of compounds that could illuminate SAR for pyranicin. With this in mind, we settled on a strategy that called for the preparation of three key subunits: pyran 2, alkynyl olefin 3, and butenolide 4. Pyran 2 would be prepared by our recently described Achamatowicz oxidation-Kishi reduction sequence,⁵ and our plans for subunit coupling involved the use of Carreira’s asymmetric alkynylation⁶ and Fu’s recently described alkyl-alkyl Suzuki coupling.⁷,

The synthesis of the tetahydropyran-containing domain commenced with commercially available furan 5⁸ (Scheme 1). Addition of dodecylmagnesium bromide produced furfuryl alcohol 6, which was subjected to the Sharpless asymmetric kinetic resolution reported by Sato^9,10 to yield the intermediate pyranone hemiacetal 7. Immediate reduction with i-Pr₃SiH in the presence of BF₃·OEt₂ gave pyran 8 as a single diastereoisomer (>20:1 by ¹H NMR analysis). This three step sequence conveniently provided the desired pyran in 33% overall yield. Hydrogenation of the enone, followed by reduction of the ketone with L-Selectride gave axial alcohol 9 in 86% yield (2 steps). Protection of the secondary alcohol as the TBS ether, removal of the benzyl ether (88% yield, 2 steps), and oxidation with Dess-Martin periodinane led to aldehyde 2 (99% yield). Reaction with known alkyne 3¹¹ under Carreira’s asymmetric alkynylation conditions (Et₃N, Zn(OTf)₂, (−)-NME, PhMe) and subsequent protection of the secondary alcohol as the TBS ether gave 10 (82% yield over 2 steps).

Scheme 1.

Synthesis of the C7–C32 pyran-containing domain.

The synthesis of butenolide 4 began with the alkylation of alkyne 11¹² with epoxide 12¹³ in the presence of BF₃·OEt₂, followed by immediate protection of the homopropargyl alcohol with TBDSPCl to give 13 in 88% yield (Scheme 2). Careful removal of the TBS ether with ethanolic PPTS (95% yield) and hydroalumination-iodination of the alkyne following Denmark’s procedure¹⁴ produced vinyl iodide 14 in 90% yield. In accord with Stille’s original report¹⁵ and Hoye’s application in the context of acetogenin synthesis,¹⁶ when 14 was subjected to Pd(0) under 50 psi CO, carbonylative lactonization occurred to yield the desired butenolide 15 in 90% yield. Oxidative removal of the PMB ether with DDQ in wet CH₂Cl₂ (82% yield) unveiled primary alcohol 16. Conversion of the alcohol to the bromide 4 was readily achieved by Hannesian’s NBS-Ph₃P protocol¹⁷ in 87% yield.

Scheme 2.

Synthesis of the butenolide domain and completion of the synthesis.

With routes to 10 and 4 secured we were positioned to examine the proposed Fu-type Suzuki coupling for subunit union (Scheme 2).¹⁸ To this end, when 10 was treated with 1.1 equivalents of freshly prepared 9-BBN in THF at room temperature, hydroboration was selective for the terminal alkene over the internal alkyne¹⁹ to yield intermediate borane 17. When this borane was exposed to bromide 4, Pd(PCy₃)₂ and K₃PO₄.H₂O smooth cross-coupling ensued to yield the desired compound 18 in 60% yield. In our hands, this coupling was robust with 20 mol% Pd(PCy₃)₂ catalyst loading and 1.2 equivalents of bromide 4. In the current example, deviation from these conditions resulted in decreased and variable yields of the desired compound. The synthesis was then completed in a straightforward fashion by removal of the silyl protecting groups with aqueous HF in acetonitrile (92% yield), and reduction of the alkyne to the hydrocarbon using Wilkinson’s catalyst in benzene-ethanol to produce pyranicin, 1 in 89% yield. The physical characteristics of synthetic material (¹H and ¹³C NMR, HRMS) matched those reported by McLaughlin.² Optical rotation data was in accord with that reported by Rein and Takahashi.^20,21

In conclusion, we have described a concise (13 steps longest linear sequence) and efficient (10% overall yield) synthesis of pyranicin. Beyond these metrics, the synthesis is modular and will prove amenable to the synthesis of a variety of analogs. This work also highlights the utility of the Achmatowicz oxidation-Kishi reduction sequence for the assembly of pyrans and the strategic potential of alkyl-alkyl Suzuki couplings in a complex setting.